NICKEL UNDER MODERATE PRESSURES. 
361 
pressure-chamber was such that the window was seven inches from the spark-gap 
the window soon became clouded and had to be cleaned after each exposure. 
This effect was still more strongly marked when the nickel electrodes were replaced by 
others of a nickel-steel containing 25 per cent, of nickel. 
The Spectrograph. —The 10 ft. 6 in. Rowland concave grating spectrograph in the 
Spectroscopic Laboratory of the Imperial College of Science and Technology was 
used for this work. The grating is mounted in the Littrow manner described by 
Eagle. # 
One of the greatest advantages of this method of mounting is that the whole 
apparatus is completely enclosed in a double-walled box, so that spurious shifts due to 
temperature variations are much less likely to give trouble than would be the case 
with the usual Rowland mounting. The importance of this feature in work involving 
the measurement of very small displacements will be sufficiently obvious. 
The method of taking the photographs was as follows :—A horizontal spark was 
employed, as it was found that in this position the wandering of the spark seldom 
caused the image to leave the slit, thus effecting a considerable economy in the time 
of exposure. The astigmatism of the grating served to give sufficient length to the 
spectrum lines. In all the plates intended for measurement, the third order spectrum 
o 
was employed, giving a linear dispersion of about 17 A.U. per millimetre. 
Each plate was exposed in three strips, the pressure spectrum being in the middle 
with a normal comparison spectrum just touching it on either side. One of the 
comparison spectra was photographed before the pressure spectrum and one after, so 
that temperature-shifts, if any, could be easily detected and allowed for. An iron arc 
spectrum was also photographed on the same plate for purposes of identification and 
for convenience in determining the dispersion-factor of the measuring apparatus. 
The region investigated for purposes of measurement was included between 
X 3450 and X 4600, the exposures ranging from 5 to 20 minutes on Lumiere ultra- 
rapid plates. The changes in the character of the lines have been investigated as far 
as X 6100 in the first order. 
3. The Measuring Apparatus. 
A Hilger measuring-machine, of the usual type in which the microscope is driven 
forward by means of a finely cut screw, was employed. The ordinary eyepiece was, 
however, replaced by one in which the cross-wires could be made to travel across 
the field of view by means of a micrometer screw. 
The arrangement of cross-wires is shown in fig. 1. Three parallel wires, a, 6 l5 b 2 
are attached to the sliding frame, the two latter being at such a distance apart that 
the image of a line of average intensity nearly fills the space between them. The 
third wire a is about four times as far away and is only used when measurements are 
VOL. CCXIV.-A. 
* 1 Astrophysical Journal,’ XXXI. (1910), p. 120. 
3 A 
